Greatest Peaks Project (2022): #13 - 2020-21 Giannis Antetokounmpo

[f4p]

So I was going to hold off on posting until I had something really substantive, but this takes precedence.

Sure, not all MVP's are equal. Steve Nash has 2, but I think we can all agree at least a little of that had to do with him being uhh...Canadian. And Moses was definitely quite uhh...non-Canadian, and still got 3 of them. Won 2 playoff series against Magic and Kareem, outplaying Kareem in both, made the Finals on a 40 win team, then seemingly dominated the playoffs for a dominant playoff team. His relentlessness seems to have had an effect on both winning and on voters.

Yeah Steve Nash won two of the only three MVPs awarded to white players from 1987-2020 because the voting bloc just loves white guys that much.

If the priority is to vote for players with an MVP and Finals MVP, what puts Moses above Willis Reed or Bill Walton (or Mikan or Pettit, for the older school voters)? Why put so much stock into Moses winning a three game series and then going “fo’ [five] fo’” on a team that had been to two of the past three finals (and probably would have won a championship in the year they missed if they had made one more basket in Game 7)?

Based on voting patterns so far, Moses will likely be admitted before I consider voting for him (see beginning of post), and that is fine, but it would be nice if the rationale given were a little more comparative. Especially if the one comparison made is that well “obviously” it was harder for Moses to win multiple MVPs than it was for white Steve Nash.

Maybe Moses should have been ranked higher than Larry Bird: after all, Moses was just so much better that he not only made up for Bird’s team winning seventeen more games and a head-to-head Finals, but he also overcame the innate voting bias toward Bird’s whiteness!

I don't think suggesting Steve Nash being white helped him a little is that out there. It was certainly implied/joked about/argued plenty at the time he was actually winning. It doesn't mean voters were walking out of Klan meetings to vote for Nash, just that it was novel and probably put a little extra pep in their steps when it came time to vote. I think it certainly works in a similar way to how Curry doesn't "look the part" and that tends to get him a little extra juice in these things. Certainly among casuals, and voters are casuals, by and large. Anyway, the main point was that MVP's aren't everything but 3 of them must mean something. Especially, when maybe the only other controversial/arguable even 2 time winner could have had a little help/media push that certainly wouldn't be applicable to Moses.

[AEnigma]

So I was going to hold off on posting until I had something really substantive, but this takes precedence.

Sure, not all MVP's are equal. Steve Nash has 2, but I think we can all agree at least a little of that had to do with him being uhh...Canadian. And Moses was definitely quite uhh...non-Canadian, and still got 3 of them. Won 2 playoff series against Magic and Kareem, outplaying Kareem in both, made the Finals on a 40 win team, then seemingly dominated the playoffs for a dominant playoff team. His relentlessness seems to have had an effect on both winning and on voters.

Yeah Steve Nash won two of the only three MVPs awarded to white players from 1987-2020 because the voting bloc just loves white guys that much.

If the priority is to vote for players with an MVP and Finals MVP, what puts Moses above Willis Reed or Bill Walton (or Mikan or Pettit, for the older school voters)? Why put so much stock into Moses winning a three game series and then going “fo’ [five] fo’” on a team that had been to two of the past three finals (and probably would have won a championship in the year they missed if they had made one more basket in Game 7)?

Based on voting patterns so far, Moses will likely be admitted before I consider voting for him (see beginning of post), and that is fine, but it would be nice if the rationale given were a little more comparative. Especially if the one comparison made is that well “obviously” it was harder for Moses to win multiple MVPs than it was for white Steve Nash.

Maybe Moses should have been ranked higher than Larry Bird: after all, Moses was just so much better that he not only made up for Bird’s team winning seventeen more games and a head-to-head Finals, but he also overcame the innate voting bias toward Bird’s whiteness!

I don't think suggesting Steve Nash being white helped him a little is that out there. It was certainly implied/joked about/argued plenty at the time he was actually winning. It doesn't mean voters were walking out of Klan meetings to vote for Nash, just that it was novel and probably put a little extra pep in their steps when it came time to vote. I think it certainly works in a similar way to how Curry doesn't "look the part" and that tends to get him a little extra juice in these things. Certainly among casuals, and voters are casuals, by and large. Anyway, the main point was that MVP's aren't everything but 3 of them must mean something. Especially, when maybe the only other controversial/arguable even 2 time winner could have had a little help/media push that certainly wouldn't be applicable to Moses.

So because lazy jokes and dumb arguments were made then (mostly by people who have always disrespected Nash), we should make them now? You talk about casuals: tell me, how do casuals feel about Nash? How many casuals prefer results over gaudy box scores? How many casuals continue to see those MVPs as robberies? How many casuals weep for stolen awards from Shaq and Kobe and Lebron (notoriously very un-casual friendly).

Nash was novel in 2005, sure, but not because he was white. He won because he joined a lottery team and revolutionised the league en route to a comfortable first seed. Want to see whether that corresponds better to MVPs than “being white”?

Did he beat Dirk in 2006 because he was whiter? I guess he lost to Dirk in 2007 because the “novelty” wore off, but seeing as he still finished second, maybe voters needed one more year to really satisfy that white itch before light-skinned Curry came in and gave them an easy out over dark-skinned Harden.

Moses won in 1979 and 1982 by virtue of playing heavy minutes and being a combined point and rebound monster. Superficial biases worked for him, not against him. Otherwise, your theory would dictate that Bird would have won comfortably in 1982.

Or maybe voters just got more casual twenty-five years later.

[AEnigma]

Okay, I am going to use this post to explore some of the thoughts I have been reading while using that exploration to touch on how I am assessing these players. It is not meant to be revelatory; I have not voted before though, so I think this would save time later on if I can knock out some of these introductory thoughts right here. The next post or two will have more specific content to this thread.

I wanted to wait for the top twelve to finish -- it was possible that someone new could have broken in, but my doubts held true -- because I think the discourse for those players is kind-of tired and stagnant. However, I think the Garnett thread had a lot of material to bridge the transition from that expected top twelve into the more nebulous remainder of the project.

Garnett’s profile is directly comparable to Duncan’s -- as more than a few brought up -- to an extent not true of any other duo. They played at the same time, they played the same position, and their overall profile is reasonably close; to some extent this could also be said of Hakeem and Duncan, but peaking a decade apart already leaves significant room for interpretation differences. So Garnett supporters tend to adopt a mentality that those who do not place them alongside each other are either being inconsistent or are ignorant of whatever relevant context would place them alongside each other. One poster even highlighted that in 2015, this project did in fact place them right next to each other -- and in 2012, Lebron was the one player in-between, so you could argue they may as well have been right next to each other. In 2015, you essentially see a swap of the Bird/Magic pair and the Duncan/Garnett pair, as well as a swap between Lebron and Bill Russell (ouch). Then in 2019, we see the duos break further: now Duncan and Bird are next to each other, and Magic and Garnett are next to each other, with Russell and now Hakeem (ouch) separating. But I will note, for the third time, Garnett is #11. In this project, he is #12, because Curry finally broke into that top 11 group of peaks and in all likelihood enshrined a new top 12, presumably buoyed by a Finals MVP won on his weakest team since 2014. Point is, Garnett at least has been mostly consistent: he has always been below this project’s top 7, and 2015 was the outlier year where he shot ahead of Russell, Bird, and Magic for the only time. (Sidenote: the usual three after Garnett and now Curry has traditionally been Walton, Oscar, and Erving in some order; with Giannis’s ascent, looks like that might permanently change too.)

This is likely where the much maligned “winning bias” comes into play. Garnett is the only player in that top twelve group who did not win multiple titles and a Finals MVP (he did neither). We do all recognise him as the best player on that 2008 Celtics team, but there too it is hard to deny that his numbers in the 2008 Finals make it feel reasonable enough that he would lose the award to Pierce, even if a Playoff MVP (or the newly introduced Conference MVP) would lean toward Garnett. The contention is then that many voters are not really analysing the data and are instead prioritising acontextual achievements.

Possibly true -- but this is not really the trap data hounds like to portray it as, and if anything, this focus on “internal consistency” is a lot more likely to reveal inconsistencies in a data-obsessive position than it is to reveal them in an achievement-focused position. Because even if it may seem arbitrary, and in fact a large number of the votes for Garnett themselves show how to some extent how that is arbitrary (Garnett: definitely worse than those eleven guys, but also meets a cutoff putting him more comfortably above Walton, Giannis, Erving, Kobe, Wade, Reed...), that type of assessment has a more immediate baseline than what the over-reliance on “impact” offers. Using a collection of all-in-one numbers to compare disparate years in disparate situations with disparate plus-minus methodologies across different player types and sometimes even internally disparate metrics (stop comparing PIPM pre- and post-1997!), to the extent that itself is done consistently either (will never stop being surprised at how unanimous the support is for Shaq in these projects), is a lot more convoluted and isolated from “player quality” than simply looking at how guys like Duncan and Magic and Bird elevated their games against the best competition even in situations where they were not receiving a lot of help (if not exactly as little as what peak Garnett had).

And to be clear, I personally would be fine throwing Garnett’s peak anywhere in that 8-12 range. But it is not a demonstrable answer that he was a more valuable postseason asset than Magic or Bird or Russell. That question is contentious even if we look at within set eras and positions; how are we going to act like it is wild for there to be a gap between him and Duncan when people are asked to start comparing his relative with perimetre stars from the 1980s? We do not need to play dumb in questioning why real achievements are more easily valued over hypothetical ones: we know that Garnett can win a title with the 2008 Celtics while past his peak but still in the tail end of his prime, maintain as a title contender while further aging along with that same team, and can at his absolute peak drag a bad team to a five-or-six-game loss against legitimate title contenders. That is a pretty wide range of possibility, even if you give him every benefit of the doubt for “era translation” and “portability”, and without any defining (small sample!) series like what we saw from Duncan in 2002 and 2003 or what we saw from Bird in 1984 and 1986 (and the 1981 conference finals).

To that point…

[AEnigma]

A Case for Robinson > Giannis

This really irks me, and is part of a broader pattern that I find particularly annoying with data obsessives.

1) Impact metrics: Robinson > Giannis. Robinson wins in 5 of the metrics, while Giannis wins in only 1. They're tied in 1 more. In the stats we don't have for each of them, Robinson is 1st while Giannis is 2nd. Robinson wins in the "true" plus minus metrics like AuPM/RAPM/PIPM/WOWY while Giannis only looks comparable or better in the box-stats.

See, this is just outright misuse. Most of this is not readily available, or at least not in equivalent form, for the playoffs in Robinson’s prime (which is his primary criticism to an extent that goes well beyond every other top player). In the regular season, PIPM is not equivalent for both, and we do not even have Giannis’s PIPM from the past three years anyway (his 2020 LEBRON is the highest in the database going back to 2010 and represents a significant jump from 2019). We do not have RAPM for Robinson’s prime.

Of course, the problem here is not just one of “playoffs versus regular season” or “what metrics are available for what purpose or in what form”. It is also treating it as something of a given that these reflect innate player quality -- but not consistently, mind you, or else we would need to start having some interesting conversations about 2003 Dirk (we will come back to that one) and 2016 Draymond and 2015/2017 Chris Paul and 2021/22 Jokic and oh right 2021 Gobert. Sure, these are reasonably correlative with player quality on a year-to-year basis, but they become dramatically less useful looking across eras and at different situations. And again, with Robinson this is mostly restricted to regular season only, which weirdly becomes something of a shield: oh, we do not have data to quantify to what extent he was “responsible” for his team getting torched by the Jazz in 1994 and 1996, so we should just continue to use his regular season as a logical jumping off point.

Sidenote: this aspect of missing data frustrated me with a lot of the Hakeem discourse. In addition to 1993 Hakeem getting moderately dismissed for happening one year prior to our access to on/off data, there are strong indicators that Hakeem likely saw a similar postseason climb in “impact” as what we saw from 2001-03 postseason Duncan, but because we do not have those numbers, we instead need to treat Hakeem at a lower-level than make any real effort to quantify them or even treat them comparatively (such as with someone like Shaq, if we take the approach that peak Duncan’s postseason values are too unreasonably high to project anyone to have).

But because so much of Robinson is tied to regular season data, people here happily fall into the trap that “giant data” = “best player/season”, even though we broadly know that is not really true (2009 Lebron versus 2012/13 Lebron versus 2016/17 Lebron as the easiest example thus far, although I do think 2016 is always a funny deliberate exclusion from these conversations because of how it inconvenient it is for the usual data narratives). I said we would come back to 2003 Dirk. 2003 Dirk dwarfs other Dirks in AuPM and (relatedly) PIPM. Now, 2003 Dirk is a bit of a lost postseason, and we know his best postseason was 2011, and we know he had high RAPM values in 2011 that more consistently outperform what we saw in 2003 (but not always, which is another little problem with just pointing to whatever data source we have on hand), so it makes it easier to say that nope 2003 was not his real peak, 2011 was. Okay, fine. What about 2006? Again I think the interrupted postseason would end up being the main distinction, and retrospectively the existence of that 2009-11 stretch is what makes people more willing to believe that 2003 was clearly pre-peak, but I think in a different history -- maybe Dirk gets majorly injured in 2010 -- that a lot of those data obsessives would be more regularly looking at 2003 Dirk. This is of course not unique to Dirk or Lebron either: Stockton saw a massive impact spike in 2001, Mookie Blaylock boasted some massive impact numbers in 1997, Mike Conley boasted some massive impact numbers in 2013… The problem with treating these numbers as objective markers of quality is that is never what they were designed to do… but that is also the only reason you would be all in on 1994 Robinson.

1994 is the season where Robinson was asked to do the most. In 1995 Avery Johnson came in and took over a fair chunk of his creation load. In 1996 the Spurs brought in Will Perdue and his “impact” went down even further (an inverse of how losing Howard Eisley made Stockton’s “impact” skyrocket for the 2001 Jazz). And it would be one thing if that were your intent: specifically taking the “most valuable” seasons. But that does not seem to be how you (or any of the other backers of 1994 Robinson) have voted, which makes me think this is more to do with a fundamental misunderstanding of what these numbers are saying. His “team fit” may not have been “ideal” in the sense you argue it is for Giannis, but it was pretty ideal for creating massive value results. As the fit gets better, his value goes down, yet you stick with the bad fit high value option (this is particularly interesting in the postseason, but we will get to that) while still trying to complain about the disadvantages of bad fit.

Resilience: ? Giannis has a reputation of being more resilient, but I’ve argued otherwise. Giannis had all-time level scoring/offensive declines in 19-20. Even if he improved in 2021 as a player, he still showed some playoff-decline, and this decline increased against better teams (he declines even after adjusting for opponent defense). I’m not sure Giannis has the true resilience advantage, and it’s certainly not much if so…

Resilience Concerns for Giannis: tl;dr 1) Massive scoring decline, particularly an efficiency decline in midrange/free throw/3 point shooting. 2) Giannis doesn't improve enough in his best series compared to Curry. 3) Despite offensive improvement in 21/22, he still declines more against better defenses than other peaks, likely because they can take away his rim attempts. 4) Personally, I don't see the defense as enough to make up for the offensive drop.

1) Scoring Decline: In literally every prime playoff, Giannis has had a scoring decline, and scoring is his best offensive trait:
in 2019 playoffs: Massive drop in scoring (-2.2 pts/75 drop, -6.8% rTS efficiency drop)
in 2020 playoffs: Drop in scoring (-2.2 pts/75 drop, -1.4% rTS efficiency drop)
in 2021 playoffs: Drop in scoring (-0.5 pts/75 drop, -3.3% rTS efficiency drop)
in 2022 playoffs: Massive drop in scoring (-1.3 pts/75 drop, -6.8% rTS efficiency drop)

There's a clear shooting decline from the free throw line, long midrange, and 3 point line, and a rim-scoring decline against tougher defenses Giannis has declined in Free Throw percentage in literally every playoff run, at an average of -9.2%

-Against bottom 10 defenses (21 Nets, 21 Hawks, 22 Bulls): -0.0 ScoreVal, -0.1 pts75, -1.5 rTS%
-Against other top 10 defenses (21 Miami, 21 Suns, not including 22 Celtics): +0.1 Scoreval, -0.2 pts/75, -4.5 rTS%
-Against all Top 10 defenses (21 Miami, 21 Suns, 22 Celtics): -1.1 ScoreVal, -0.5 pts/75, -6.9 rTS%
So Giannis has shows small decline in volume and efficiency against bad defenses, a large decline in efficiency against top defenses even if we don't include the Celtics, and an all-time decline in efficiency if we don't filter out the Celtics… Question for people: Are we sure we're not biased for Giannis' resilience just because of a single playoff series vs Suns?

I like how you excised Giannis’s first-team all-NBA season in 2018 because it was not “prime”. Convenient, considering how well he performed against the best defence in the league.

Anyway, the question for Giannis has never been, “Is he a top tier resilient playoff scorer?” The question is something more along the lines of, “Is his scoring resilient enough as a secondary skill to his defence?” And that is a difficult question to answer, but by comparison with David Robinson, a more thorough analysis should pretty comfortably reflect that the answer is yes. It is not enough to just look at relative dropoffs -- you should know this, but as a refresher, if someone is starting from a higher base, then any dropoff could still leave them well ahead of their point of comparison. Giannis is a better scorer than David Robinson (and Kevin Garnett). Outright. His volume is a lot higher and his volume maintains better. The dropoff in efficiency is a little more debatable -- tied to the idea that Giannis’s scoring game might be baseline less suited for the playoffs, he can see a bit more of consistent efficiency dropoff than Robinson. But against good defences, yeah, Giannis’s efficiency holds up at least as well, if not better considering his higher starting point.

Not going to spend as much time on this because others have pointed out flaws here, but as a direct comparison with Robinson, those aspects have been touched on much less.

I'd argue much of Robinson's perceived lack of resilience is actually just from being on a worse team, which allowed opponent defenses to focus on him on offense, and forced him to run out of motor making up work on the defensive end. Why else would his resilience "suddenly" improve when the Spurs got better teammates in '99 onward? Giannis on the other hand had a great fitting offensive and defensive team in 21. 3-point Spacers at all the positions and even at the 5. On defense, he had an All-nba point-of-attack/perimeter defender and a twin tower to help protect the rim

First off, no his resilience did not improve with Duncan. Lol. You did a little trick here like what you did with 2018 Giannis by just throwing out a year that is inconvenient to the narrative you want to push. It is true that Robinson was less affected as his scoring volume went down, and yes, that is a product of being a clear second option rather than a first option (or in the case of 1998, a co-first). But we are not comparing Giannis as a second option to Tim Duncan with old David Robinson as a second option to Tim Duncan. That is an impossible comparison, but in any case, Giannis is a much more prominent first option than David Robinson ever was.

Second, yeah, the Bucks strictly speaking have 3-point spacers, but for the most part they are not really reliable ones in the postseason. That is part of the problem. Going to jump to a different comment here.

The greater scoring dip against better defenses is because better defenses can better take away his rim attempts in the half court. It's worth noting that the Bucks as a whole also show a greater decline against better teams than other Greatest Peak players' teams, so I tend to think Giannis' resilience disadvantage does impact team results.

You blame this on Giannis, as if he individually controls the ability for his teammates to make shots. Like, this might be the most frustratingly disingenuous framing yet. When Curry is doubled, and his efficiency dips, and his teammates take advantage, that is praiseworthy. Curry is so brilliant. But when multiple defenders collapse in on Giannis or commit to walling him off, and his efficiency dips, and his teammates fail to take advantage… well, that is also Giannis’s fault. You have watched those series, right? This is not a case where Giannis gets picked up in single coverage and just dumps it off to a guarded teammate. His teammates simply have failed to make shots. Maybe you blame that on them, maybe you blame that on the scheme, but you do not blame it on the guy drawing the full focus of the opposing defence.

Anyway, back to Robinson. The main point is that yes, it is tough to be a first option. And we can look at Robinson’s playoff history for a demonstration: from 1990-93, when he had some reasonable offensive support and was only asked to score around 24 points per 75 possessions, he saw much less of a dropoff than when he was shouldering more of the volume. That is usually how it goes, and the idea of struggling as a heavy load first option is not some unique excuse. Pretty much every star benefits from being focused on less. But what you are trying to do is suggest that because Robinson had worse teammates he was de facto focused on more. No. Robinson, even at the peak of his offensive responsibility, was neither doing as much on offence as Giannis nor was prone to the type of defensive attention that teams can throw at star players today. There is this trend to blindly talk about spacing as if that is the only real consideration. Modern defences have evolved along with modern offences. You make claims like:

Time machine: I see Robinson doing well in today's era, but like I've mentioned before, I'm worried about Giannis in an earlier era. His defense would do great of course. But on offense, so much of it relies on spacing and loose dribbling rules.

… while ignoring how Robinson would adapt to modern schemes on either end. Okay, sure, Giannis gets better spacing and looser dribbling rules (do you want Robinson to dribble more?). He also gets to deal with two or three players contesting him every time he drives in the half-court. This tired myth about the 1990s being just so much more impossibly difficult needs to die. In the worst case, Giannis is a supercharged Shawn Kemp, and in Robinson’s best case as a scorer, he is a less effective Joel Embiid. And on defence?

Defense: Robinson > Giannis, at least personally. Robinson's absolutely better rim protector, and a fantastic overall help defender. With Duncan, Robinson helped lead the greatest defensive dynasty of the modern era. I see this as part of the reason BPM underrates Robinson.

And what would Giannis do with Tim Duncan? How would Robinson adapt to modern offences when even 1990s offences like the 1991 Warriors and the 1994/1996 Jazz comfortably neutralised his defensive impact? This is not analysis. (By the way, BPM if anything may as well be designed to accentuate Robinson’s defensive impact. He is the career leader because his defence shows up on the boxscore in a way less true of pretty much every other historical player.)

The problem here, and maybe just as a throughline in the entire argument, is that you seem to be taking this “data” that to you says Robinson was “better” and just adapting scenarios to fit that. Robinson is a scalable data giant, ergo he fares well in the modern era. Robinson had bad postseason results despite a massive regular season data footprint, and with Tim Duncan he put up a strong postseason data footprint, ergo we should assume that his prime postseason failings were primarily no real fault of his own. You can set your standards however you see fit -- seems like “biggest impact to team, with edits” -- but do not be surprised when other people take issue with that framing or outright reject it, because none of that is real production. And if you want to make a production argument, and if you want to convince people with that production argument, you need to be a little more analytical than focusing on extreme negative framings for all contenders and focusing on extreme positive framings for the guy you want admitted.

You are very polite, but it is frustrating to read long posts with such weird approaches to discerning which player is “better” that also do not seem to spend much time truly trying to understand how other people are making their comparisons.

[AEnigma]

Final post in this chain.

Consistency is something I have seen harped on repeatedly, in this project and in most prior projects. Often when it is brought up, I think it relies too sincerely on the questioner’s own feelings of consistency (e.g. Garnett should be right by Duncan) rather than on how other people might pursue internal consistency. And there is value in establishing that dialogue, or perhaps using it to some rhetorical effect (when I ask what distinguishes Moses, I have suspicions of what I think are the likely answers, but I want them established first), but too frequently I think the result is minimally constructive. To preclude that a bit, here is a partial summary of my approach:

I think Lebron is a clear outlier. His impact on his teams, across different teams, across different roster constructions, across different roles, to me looks wholly unprecedented, in the data era and in the most talented and schematically advanced the league has ever been. I say he is an outlier in part because I think bigs historically have more absolute value than any comparatively equivalent perimetre player. I say absolute value because I think this can in some situations hamper how bigs look in terms of their relative value: Michael Jordan is more of an outlier than Hakeem or Kareem or Russell, even if I suspect that Hakeem or Kareem or Russell could potentially offer more outright value. And part of why I think that is because I think the defence of bigs is all-encompassing (which has made me a lot lower on Shaq, to an extent that has carried into Jokic too). Even bench bigs can usually give you a tolerable enough defensive floor by virtue of just being tall enough to deter shots at the rim (of course, if you lack adequate bench bigs, your “value” can quickly get juiced). But Lebron I think goes beyond even that positional outlier and simply stands as an absolute outlier. I am not sure you can ever really reduce his real value to a given team past a still extremely high point.

That said, after Lebron, and I guess to some extent Jordan (I do not have him locked in as my #2 peak, but it is often excessively antagonistic to argue for Hakeem as a contemporary, and arguments for Kareem or Russell or Wilt rely too much on the idea of standing out more in their era), I do prioritise bigs over perimetre players… to a point (which we should reach soon). Because to the extent I can value bigs more than perimetre players, there does need to be a cutoff where the gap between the top tier bigs and the next tier leaves room for perimetre names like Kobe and Kawhi and the like. But once we go through some of those names, I will probably go back into a consistent pattern of voting for bigs which -- judging by past results -- might take me to the end of the project. Part of what holds that next tier back is the increased idea of replaceability: if the tier 1 bigs are almost impossible to replace, and perimetre players already are starting from a point where they are innately tougher to replace, then the second tier of bigs is where I start to think that the best perimetre names are too reliable in the postseason to continue being ignored.

On that note:

1. 2020-21 Giannis (or 2022 if necessary)
2. 1976-77 Bill Walton
3. 2007-08 Kobe Bryant

Giannis and Walton close out the top bigs for me, and I can go in either order. Currently leaning Giannis just because he has proved it in a better league and did it with a playstyle that I think lends itself more naturally to modern team-building, but I think Walton was a brilliant defender (better intuitive understanding of the game than David Robinson as the next main big) and close to ideal team hub on offence (again, to an extent well beyond David Robinson, even though yes David Robinson has more to offer as a scorer). Relative to his era, Walton is of course a clearer standout, but that is true for most old players. And I do value proving your ability to bring a team to a title. Hypotheticals are nice, but I know Bill Walton could win a title with the 1977 Trail Blazers, I think there are substantial indicators he would have repeated in 1978, and there is nothing in David Robinson’s history that makes me similarly confident he could do the same (although to be clear, I would give him good odds).

Kobe I think is an interesting profile. In terms of raw value to his team, he never really hit the same heights as Dwyane Wade. But I think Wade is a lot more innately limiting in building a championship team; deserves his due for the 2006 playoff run, certainly, but I personally am not just gunning for the best individual playoff runs. Although I do make note of them.

Also heavily considering Kawhi, but 2017 lacks playmaking, and I am not sure if I feel strongly enough about 2019 or 2021 to go with those years this early. Unibro makes a strong Davis case; I could be swayed on that, but I am more wary of chasing hot streaks, and 2018 was not so incredible that I have total faith in Davis’s ability to approach that level again (it was an incredible postseason, though, no doubt there).

[70sFan]

I missed a few threads, but I hope I will contribute when I come back to home.

[DraymondGold]

This is likely where the much maligned “winning bias” comes into play. Garnett is the only player in that top twelve group who did not win multiple titles and a Finals MVP (he did neither). We do all recognise him as the best player on that 2008 Celtics team, but there too it is hard to deny that his numbers in the 2008 Finals make it feel reasonable enough that he would lose the award to Pierce, even if a Playoff MVP (or the newly introduced Conference MVP) would lean toward Garnett. The contention is then that many voters are not really analysing the data and are instead prioritising acontextual achievements.

Possibly true -- but this is not really the trap data hounds like to portray it as, and if anything, this focus on “internal consistency” is a lot more likely to reveal inconsistencies in a data-obsessive position than it is to reveal them in an achievement-focused position. Because even if it may seem arbitrary, and in fact a large number of the votes for Garnett themselves show how to some extent how that is arbitrary (Garnett: definitely worse than those eleven guys, but also meets a cutoff putting him more comfortably above Walton, Giannis, Erving, Kobe, Wade, Reed...), that type of assessment has a more immediate baseline than what the over-reliance on “impact” offers. Using a collection of all-in-one numbers to compare disparate years in disparate situations with disparate plus-minus methodologies across different player types and sometimes even internally disparate metrics (stop comparing PIPM pre- and post-1997!), to the extent that itself is done consistently either (will never stop being surprised at how unanimous the support is for Shaq in these projects), is a lot more convoluted and isolated from “player quality” than simply looking at how guys like Duncan and Magic and Bird elevated their games against the best competition even in situations where they were not receiving a lot of help (if not exactly as little as what peak Garnett had).
…
We do not need to play dumb in questioning why real achievements are more easily valued over hypothetical ones: we know that Garnett can win a title with the 2008 Celtics while past his peak but still in the tail end of his prime, maintain as a title contender while further aging along with that same team, and can at his absolute peak drag a bad team to a five-or-six-game loss against legitimate title contenders. That is a pretty wide range of possibility, even if you give him every benefit of the doubt for “era translation” and “portability”, and without any defining (small sample!) series like what we saw from Duncan in 2002 and 2003 or what we saw from Bird in 1984 and 1986 (and the 1981 conference finals).

This seems like the crux of our disagreement (and it pops up multiple times in the rest of your post), so let’s start here.

Consistency of the Data-driven approach: Your primary argument that a data-focus is inconsistent replies on three things (at least as I understand it):
1) You suggest a data focus isn’t an “immediate baseline”. You call it a “hypothetical achievement”, compared to the achievement-focus (e.g. winning a championship) which is a “real achievement”
2) You suggest a data focus is “convoluted,” while an achievement-focus is less complex.
3) You suggest a data focus lacks sufficient context and analysis (e.g. when comparing players in different eras, different situations, potentially with different teammates/opponents).

For 1: I’m not sure why being an “immediate baseline” should necessarily be proof that an achievement focus is consistent, while a data focus isn’t. It’s not true that having a good impact metric is a “hypothetical achievement”. It is a true fact that Giannis was an NBA champion in 2021. Likewise, it is a true fact that Giannis scored 28.1 ppg in the regular season. Likewise, it is a true fact that Giannis scored a +5.03 Goldstein RAPM. There’s nothing hypothetical about any of this.
Takeaway: Data not necessarily inconsistent.

For 2: You talk about increased complexity of data over “achievements” like this makes data inherently inconsistent. I’m not sure this is true. Things can be consistent and complex, or simple and inconsistent.
Takeaway: Data not necessarily inconsistent.

For 3: This is the most compelling point. It’s true that just looking at stats without context can be misleading for the reasons you suggest. But here’s the problem: this is a bit of a straw-man. This implies that everyone who used data in the past threads didn’t provide context.

But we did! I provided plenty of context. I compared eras, I analyzed the potential biases of different stats, I compared the team’s fits, I compared performance against better defenses, I compared their different skillsets… all of the things you said was missing from the data-driven approach. Now, you may not agree with someone’s evaluation of the context. But the fact that you disagree with someones interpretation of the context =/= no context was provided.

It’s worth noting — many of the times I provided context, you used this as evidence that I was biased: you say I was “throwing out a year that is inconvenient to the narrative [I] want to push
or that the metrics aren’t perfect.”, when really I was providing context for Robinson.

Similarly, you suggest that many times knowledgeable analysts provide context to the data, they’re actually just showing inconsistency. For example, you suggest that it’s inconsistent that the data-driven approach considers 2010/2011 Dirk > 2006/2003 Dirk. You make simalar arguments with mentions of Draymond/Paul/Gobert.
But… taking 2011 Dirk >2006/2003 Dirk is the very thing that context/film-analysis would suggest! Taking Draymond/Paul/Gobert lower is the very thing that context would suggest! It’s also what an achievement-approach would suggest, which is the one you’re a proponent of.

So… from your perspective, either data-driven approach either doesn’t consider context but should (which isn’t true) or they do consider context which shows inconsistency with data. This logic seems a bit inconsistent, no? Either we should consider context or we shouldn’t. In truth, we do. But just because you personally disagree with someone’s analysis of context doesn’t make it invalid or inconsistent.
Takeaway: Data not necessarily inconsistent.

Inconsistency of the “Achievement”-approach: Instead, I’d argue taking an “achievement focus” opens you up to more potential biases and inconsistencies. People have discussed “winning bias”, which you mentioned too, where the results of the team against a certain opposition are conflated with the greatness of the individual. I’ve also raised the idea of a “scoring bias” (overvaluing scoring and undervaluing playmaking), and “athleticism bias” (overrating athleticism and underrating skill/BBIQ), and “big moment bias” (over-valuing a single moment while undervaluing the larger context).

An achievement-based approach would open you up to all of these. Data doesn’t care whether you won the championship or lost… it just tells you how much you helped your team by. Data doesn’t care about whether your impact comes from scoring or passing. Data doesn’t care whether your greatness comes from athleticism, skill, or BBIQ. Data doesn’t care about a single memorable highlight which (in the vast majority of cases) doesn’t actually change who helps their team win more.

Could an achievement-based approach lead you astray? Let’s try an experiment. Let’s do an achievement-based approach for Garnett. You suggest that the “pretty wise range of possibility” in Garnett’s team results suggest a wider range of possibility compared to the consistency of “guys like Duncan and Magic and Bird [who] elevated their games against the best competition.”
Well, if team results are what we’re going by, that would suggest Garnett is better in 2008 vs 2003/2004, right? I’d be hard pressed to find anyone on this board who would agree with that.

So how do we know that Garnett was actually better in 04, using the “achievement-based” approach? Well, we’d have to look at context…. the same context people using the data-driven approach are providing. Except where just looking at “achievements” [of the team] gives us no way to isolate how much the individual helped the team, just looking at the data gives us a very precise ability to isolate how much the individual helped the team.

In sum: both methods need context. The data-driven approach gives us an much better insight into how much the individual was actually contributing to wins, while avoiding biases like winning bias, scoring bias, athleticism bias, and big moment bias. An achievement-based approach… doesn’t tell us how much the individual helped their team win and opens us up to those biases far more.

[f4p]

So I was going to hold off on posting until I had something really substantive, but this takes precedence.



Yeah Steve Nash won two of the only three MVPs awarded to white players from 1987-2020 because the voting bloc just loves white guys that much.

If the priority is to vote for players with an MVP and Finals MVP, what puts Moses above Willis Reed or Bill Walton (or Mikan or Pettit, for the older school voters)? Why put so much stock into Moses winning a three game series and then going “fo’ [five] fo’” on a team that had been to two of the past three finals (and probably would have won a championship in the year they missed if they had made one more basket in Game 7)?

Based on voting patterns so far, Moses will likely be admitted before I consider voting for him (see beginning of post), and that is fine, but it would be nice if the rationale given were a little more comparative. Especially if the one comparison made is that well “obviously” it was harder for Moses to win multiple MVPs than it was for white Steve Nash.

Maybe Moses should have been ranked higher than Larry Bird: after all, Moses was just so much better that he not only made up for Bird’s team winning seventeen more games and a head-to-head Finals, but he also overcame the innate voting bias toward Bird’s whiteness!

I don't think suggesting Steve Nash being white helped him a little is that out there. It was certainly implied/joked about/argued plenty at the time he was actually winning. It doesn't mean voters were walking out of Klan meetings to vote for Nash, just that it was novel and probably put a little extra pep in their steps when it came time to vote. I think it certainly works in a similar way to how Curry doesn't "look the part" and that tends to get him a little extra juice in these things. Certainly among casuals, and voters are casuals, by and large. Anyway, the main point was that MVP's aren't everything but 3 of them must mean something. Especially, when maybe the only other controversial/arguable even 2 time winner could have had a little help/media push that certainly wouldn't be applicable to Moses.

So because lazy jokes and dumb arguments were made then (mostly by people who have always disrespected Nash), we should make them now?

i'm not sure why you are getting so upset over a fairly off-handed comment, and one that isn't exactly unique to myself. my first google result led to a Quora response ( https://www.quora.com/Would-Steve-Nash-have-won-two-NBA-MVP-awards-had-he-not-been-white) by someone with "Senior Moderator at RealGM" in their description (6 years ago, maybe they aren't anymore). This doesn't seem like a lazy/dumb response:

So, this question bugs me, but not in the sense that I think the question should get down voted. It bugs me because I've been defending Nash against people who simply don't understand his impact for years and years and I feel like this question is likely posed by someone like this...but frankly it's not that crazy of a question to ask.

I'll start off by saying this:

If you look at team offense and regression data, what Steve Nash did in that era is off the charts. Talk to the vast majority of basketball fans and ask them if Nash might have been the best offensive player in history, and they'll call you crazy, an idiot, or a homer, or possibly all 3. But he deserves to be in that conversation none the less and any time you see someone whose been around the block like me say that, understand that we're about as surprised to say this as you are to hear it. I didn't think it was possible for someone who looked like Nash to be anywhere near that good.

Let me repeat that last part:

I didn't think it was possible for someone who looked like Nash to be anywhere near that good.

I don't think anyone did. And while that held Nash back for years and years, when Nash finally broke through, it mad him stand out all the more. When Nash became THE story in the NBA, his appearance said much of what the people "heard", and it was that amplitude of narrative that Nash drove writers to think on him further.

Let me now make an alternative question:

Would Steve Nash have won two MVPs had the (mostly black) NBA players voted instead of the (mostly white) writers?

I'll say straight out, Nash wouldn't have won either... for largely the same reason that players sided with James Harden over Steph Curry for best player last year. If you wanted an explanation based on painting a caricature of both sides:

Writers become attached to the story. Players become attached to the glamour.

And if you think those two are the same, well, most times they are, but not always.

Because writers have such a fetish for an award telling a story, anything that contributes to the salience of the story attached with an award winner has to be considered as something that seriously made a difference. And we cannot deny that Nash race made him pop out of the picture.

However, it has to be noted that Nash was only in the picture because of a perfect storm of other things, and I don't think that really needs explaining. Nash wasn't seen as an MVP candidate in previous years despite having similar volume state despite being white, so obviously big narratives took hold distinct from that.

My feeling is that if being white helped him win the MVPs, it only helped a little. That little is not nothing - it's enough that my answer to this question in the end is just:

Maybe.

You talk about casuals: tell me, how do casuals feel about Nash? How many casuals prefer results over gaudy box scores? How many casuals continue to see those MVPs as robberies? How many casuals weep for stolen awards from Shaq and Kobe and Lebron (notoriously very un-casual friendly).

we're 15 years on at this point. casuals loved westbrook in 2017 and he will probably continue to fade in the casual mind as the years go by and we are removed from "but muh triple double" mania. removed from the moment, nash's mvp's look weird in retrospect. especially because arguably his biggest support comes from the advanced stats we have now, which makes his mvp dominance with pedestrian box stats all the more peculiar, as the idea of a 15 ppg mvp would have seemed crazy before he got it.

Nash was novel in 2005, sure, but not because he was white. He won because he joined a lottery team and revolutionised the league en route to a comfortable first seed. Want to see whether that corresponds better to MVPs than “being white”?

it's less that he won in 2005. kobe and kg were on bad teams, the spurs were just being the spurs, dirk wasn't quite there in the public's mind, shaq had the "impact" narrative like nash but it was nothing like his best seasons, lebron's team wasn't good. okay, so he gets one. but you say "that sure was a weird year" and go back to traditional voting. but it looks really weird in retrospect to see steve nash with as many mvp's as shaq and hakeem combined, with 15/11 and 18/10 seasons, one for a 54 win team. and damn near a 3rd one in 2007, which would have been a travesty for him to have 3. it's the cumulative effect from a historical perspective, the double narrative benefit in back to back years, that makes people less than sure that it was without a side reason.

Moses won in 1979 and 1982 by virtue of playing heavy minutes and being a combined point and rebound monster. Superficial biases worked for him, not against him. Otherwise, your theory would dictate that Bird would have won comfortably in 1982.

Or maybe voters just got more casual twenty-five years later.

yes, superficial basketball biases. like we've seen plenty of times before.

[DraymondGold]

A Case for Robinson > Giannis

This really irks me, and is part of a broader pattern that I find particularly annoying with data obsessives.

You are very polite, but it is frustrating to read long posts with such weird approaches to discerning which player is “better” that also do not seem to spend much time truly trying to understand how other people are making their comparisons.

. Sorry to hear you're frustrated! Just to start: if you're frustrated by someone arguing for one player over another... you may be frustrated quite often around here. That's what we do. We discuss and we debate. And we don't always agree. Are you really so surprised by this?

You say that I should "spend much time truly trying to understand how other people are making their comparisons." Is this not a two-way street? The burden's just on me and on this side, and not on the other side? There's a pretty big difference between disagreeing with an evaluation and saying they are biased and aren't trying to understand a different perspective.

Data discussion:

1) Impact metrics: Robinson > Giannis. Robinson wins in 5 of the metrics, while Giannis wins in only 1. They're tied in 1 more. In the stats we don't have for each of them, Robinson is 1st while Giannis is 2nd. Robinson wins in the "true" plus minus metrics like AuPM/RAPM/PIPM/WOWY while Giannis only looks comparable or better in the box-stats.

See, this is just outright misuse. Most of this is not readily available, or at least not in equivalent form, for the playoffs in Robinson’s prime (which is his primary criticism to an extent that goes well beyond every other top player). In the regular season, PIPM is not equivalent for both, and we do not even have Giannis’s PIPM from the past three years anyway (his 2020 LEBRON is the highest in the database going back to 2010 and represents a significant jump from 2019). We do not have RAPM for Robinson’s prime.

This gets back to Point 3 in the previous post. This is how we do statistical analysis... we look at the data, then apply context/film analysis, then come to conclusions. Is started with the data. Then applied context. Of course if you just quote the data-section and ignore the context, it looks like there isn't context...

That said, your point that we don't have full data for Robinson is absolutely true! And you'll notice how often I say just that when discussing older players. I reference the limits of data for older players in that very post

A small correction: it’s not true that PIPM is not available for Giannis for the last 3 years. We have PIPM numbers for both regular season and playoffs in 2020, and we have partial PIPM numbers in 2021.
It's also note quite true that we don't have RAPM for Robinson's entire prime. We have full RAPM numbers for him from age 31, 32, and 33, and partial RAPM numbers when he was younger. I'd consider that his prime, though not peak. If 30's the cutoff for prime, then Jordan was not in his prime in 96/97/98 and LeBron wasn't in his prime in 16/17/18. If you really don't think that's prime... well, let's just agree to disagree haha

Of course, the problem here is not just one of “playoffs versus regular season” or “what metrics are available for what purpose or in what form”. It is also treating it as something of a given that these reflect innate player quality -- but not consistently, mind you, or else we would need to start having some interesting conversations about 2003 Dirk (we will come back to that one) and 2016 Draymond and 2015/2017 Chris Paul and 2021/22 Jokic and oh right 2021 Gobert.

You suggest stats don't reflect innate player quality. Perhaps you mean what I've been calling "goodness"? Because by definition they measure value. You've called data driven approach "hypothetical", but it's just as hypothetical as averaging 30 ppg is. To me and to most people, you can count things, and when you get to the final number, you can say that that factually happened. That's not hypothetical.

But yes, I absolutely agree, context needs to be provided from there! Value in a given role is not the same as goodness in any role, and we need context to decipher this. A player can be super valuable in a niche role/fit, but not as good as their value suggests if forced to play in another situation. This is the kind of context you accuse me of not applying... but it's also the same context that I do apply to understand why someone like Gobert shouldn't be the 13th GOAT Peak.

I'm also not sure the data does indeed support all the players you mentioned being in this tier. To me, that misrepresents the data.
Paul: I provided a statistical comparison with Chris Paul in a previous thread which showed across the majority of metrics, he’s a hair below the modern players we’re considering. So it’s perfectly consistent to not include Paul at #13 in a data-driven peak list… when the data doesn’t support him being #13 in a data-driven peak list.

Jokic: You suggest that I’m hypocritical for not discussing Jokic over Giannis. Except… I have. In the very post you’re quoting, I said that Jokic is in the same tier statistically as Giannis, and that I would personally prefer peak Jokic to Giannis. This seems like a bit of a straw-man, no?

Gobert: To my knowledge, Gobert is also lower across the spectrum of metrics, so in a data-driven list, it would also be consistent to have him lower. Perhaps I’m missing some metrics you’re referring to?

Draymond: Here, we face two issues.
First, sample size. 2016 Draymond shoots to the Top Tier of plenty of stats. But other Draymond years don’t, and the drop-off is far larger than the drop-off for other top single-year peaks we’re considering. I’ve argued that we should weigh single-year peaks heavily (particularly when there are contextual factors that might explain a drop on either side), but that we should still consider ~3-year peaks to get a larger, more consistent sample.
Second, is how to incorporate context into impact metrics. 2016 Draymond provides a ton of value, but when applying context, basically everyone agrees it's from having an outlier shooting season and outlier fit, which doesn't suggest how truly good he'd be in other roles. See above / previous post for more on how to include context in data.

Sidenotes:

Sidenote: this aspect of missing data frustrated me with a lot of the Hakeem discourse. In addition to 1993 Hakeem getting moderately dismissed for happening one year prior to our access to on/off data, there are strong indicators that Hakeem likely saw a similar postseason climb in “impact” as what we saw from 2001-03 postseason Duncan, but because we do not have those numbers, we instead need to treat Hakeem at a lower-level than make an real effort to quantify them or even treat them comparatively (such as with someone like Shaq, if we take the approach that peak Duncan’s postseason values are too unreasonably high to project anyone to have).

Not sure if this is directed at me or the rest of the posters…? If it’s at me, this also seems like a bit of a straw man. You can find quotes of me saying that I’d be open to taking 93 Hakeem and that 93 Hakeem did perform better than the stats we have… I just focused on 94 Hakeem because I didn’t convince anyone else to teak 93 > 94. My overall argument was: A) 93 Hakeem has greater overall impact largely from a defensive improvement, but is worse if we consider context (his lack of passing ability and lack of good fit alongside teammates is particularly concerning). B) 94 has less overall impact, but slightly improved in his weakest areas. C) Both A and B have big enough flaws that I would have him lower than he got voted in.

Sidenote: in your other post, you also complained about people having Shaq too high. I think we tend to agree on Shaq too.

Giannis / Robinson Discussion:

I like how you excised Giannis’s first-team all-NBA season in 2018 because it was not “prime”. Convenient, considering how well he performed against the best defence in the league.

This seems somewhat testy in tone for what is yet another straw man. I didn’t mention 2018 Giannis because, at least to my knowledge, not a single other poster even mentioned 2018 Giannis, much less were arguing that his 2018 performance should be indicative of his 2021 impact. Look, I'm happy to have a discussion about 2018 Giannis, but when you say something I disagree with (which is clearly explainable from the context of the rest of the discussion), my immediate instinct isn't to accuse you of intentional omission or deception. I'd strongly suggest you avoid such accusations in the future.

But because so much of Robinson is tied to regular season data, people here happily fall into the trap that “giant data” = “best player/season”, even though we broadly know that is not really true (2009 Lebron versus 2012/13 Lebron versus 2016/17 Lebron as the easiest example thus far, although I do think 2016 is always a funny deliberate exclusion from these conversations because of how it inconvenient it is for the usual data narratives).

The recipe isn't quite "giant data" = "best player".

It's "better data" + "better context that suggests the data isn't too biased towards the statistical winner / against the loser" = "better player".

2009 has better impact, but the context/film analysis is worse. That's why people who support the data-driven approach favor the older LeBron. We're applying the very context that you suggest we're missing.

Anyway, the question for Giannis has never been, “Is he a top tier resilient playoff scorer?” The question is something more along the lines of, “Is his scoring resilient enough as a secondary skill to his defence?” And that is a difficult question to answer, but by comparison with David Robinson, a more thorough analysis should pretty comfortably reflect that the answer is yes. It is not enough to just look at relative dropoffs -- you should know this, but as a refresher, if someone is starting from a higher base, then any dropoff could still leave them well ahead of their point of comparison. Giannis is a better scorer than David Robinson (and Kevin Garnett). Outright.

Of course it's not enough to look just at relative drop-offs, and I never meant to suggest it was. I actually suggest otherwise at some point in the past few threads.

It's great than Giannis is the better scorer. Outright.

But what if Robinson/Garnett are the better playmakers or off-ball players or defenders? Well, then the primary advantage Giannis has over them is in scoring.
And what if Giannis' scoring decline is large against good opponents in the playoffs? Well, then his biggest advantage over Robinson/Garnett declines.

So if Giannis doesn't shrink the defensive/playmaking/off-ball/etc. gap in the playoffs (which has been debated but certainly not proven), then Giannis might just end up being less resilient and less good overall than Robinson/Garnett. Outright.

I'd argue much of Robinson's perceived lack of resilience is actually just from being on a worse team, which allowed opponent defenses to focus on him on offense, and forced him to run out of motor making up work on the defensive end. Why else would his resilience "suddenly" improve when the Spurs got better teammates in '99 onward? Giannis on the other hand had a great fitting offensive and defensive team in 21. 3-point Spacers at all the positions and even at the 5. On defense, he had an All-nba point-of-attack/perimeter defender and a twin tower to help protect the rim

First off, no his resilience did not improve with Duncan. Lol. You did a little trick here like what you did with 2018 Giannis by just throwing out a year that is inconvenient to the narrative you want to push. It is true that Robinson was less affected as his scoring volume went down, and yes, that is a product of being a clear second option rather than a first option (or in the case of 1998, a co-first). But we are not comparing Giannis as a second option to Tim Duncan with old David Robinson as a second option to Tim Duncan. That is an impossible comparison, but in any case, Giannis is a much more prominent first option than David Robinson ever was.

Second, yeah, the Bucks strictly speaking have 3-point spacers, but for the most part they are not really reliable ones in the postseason. That is part of the problem. Going to jump to a different comment here.

The greater scoring dip against better defenses is because better defenses can better take away his rim attempts in the half court. It's worth noting that the Bucks as a whole also show a greater decline against better teams than other Greatest Peak players' teams, so I tend to think Giannis' resilience disadvantage does impact team results.

First off, once again this seems like a straw man / misrepresentation: I didn't "throw out a year" to intentionally omit for Giannis, as I mentioned above, nor is this just "a year" for Robinson. I said "1999 onward", which to me suggests multiple years.

And again with the resilience/Duncan comment, it seems like you missed the point I was making. Perhaps I could have been clearer?

Let me say unequivocally: Of course Robinson's resilience didn't magically get better with Duncan! So... why did he stop dropping in the playoffs? Well, he fit better as a second option. There's nothin magical about rookie Timmy.. the reason Robinson was better was because he was playing a role he fit more. And while it's of course true we can't see Giannis play with Duncan, we can infer how he might play as a second option based on his actual skillset and how he meshes with other offensive players on his team.

Let me summarize my argument, since that might be more productive than over-debating points that seem like straw-man arguments of mine.

Yes, Giannis is the better scorer and better offensive first option, which is a point for Giannis > Robinson.
But: Robinson may have other offensive skills over Giannis, so it's not a strong point for Giannis.
But: neither are great offensive first options compared to the rest of the competition, so it's again not a strong point for Giannis.
But: Robinson is the better defender. Point for Robinson
But: Robinson is the far better offensive second option /offensive costar than Giannis. Point for Robinson.
Thus: Robinson > Giannis.

The greater scoring dip against better defenses is because better defenses can better take away his rim attempts in the half court. It's worth noting that the Bucks as a whole also show a greater decline against better teams than other Greatest Peak players' teams, so I tend to think Giannis' resilience disadvantage does impact team results.

You blame this on Giannis, as if he individually controls the ability for his teammates to make shots. Like, this might be the most frustratingly disingenuous framing yet. When Curry is doubled, and his efficiency dips, and his teammates take advantage, that is praiseworthy. Curry is so brilliant. But when multiple defenders collapse in on Giannis or commit to walling him off, and his efficiency dips, and his teammates fail to take advantage… well, that is also Giannis’s fault. You have watched those series, right? This is not a case where Giannis gets picked up in single coverage and just dumps it off to a guarded teammate. His teammates simply have failed to make shots. Maybe you blame that on them, maybe you blame that on the scheme, but you do not blame it on the guy drawing the full focus of the opposing defence.

[/quote] This again seems like somewhat of a straw-man.

The whole concern is whether Giannis' poor playmaking is the thing that's limiting his teammates' shooting. It's great that both Curry and Giannis draw doubles. Legitimately, it's great! But I showed in film analysis that these double teams on Curry led to easier shots for his teammates, e.g. off of Curry's superior passing or superior off-ball movement.

If Giannis draws the double with the ball but struggles to pass... then his team might not get the more efficient shot, which is the whole point of drawing the double team! If Giannis draws the double without the ball, but his off-ball movement isn't in a way that leads to easier shots for his teammates... then the double team doesn't really help the offense.

Like I say in that very post, these are concerns, not proven. It's possible the fault lines 100% with Giannis' teammates. But these arguments also haven't been disproven in any in-depth film analysis either. These are all things I say in my post / in previous posts that I link to in that post. You mention you're quite frustrated. Well.. I'm also frustrated too, since my points keep getting straw-manned!

..........

Rather than rehash the other points, like the time machine (no, I don't expect Robinson to lose value by gaining more forgiving dribbling rules... yes, Robinson is a better defensive player than Embiid), I think it's time I address tone.

When I think I’ve misunderstood another poster’s meaning or intention, I simply ask. When I think I understand another poster that I disagree with, I simply explain why I disagree.

To my best effort, I never straw-man another poster, belittle them, or accuse them of being deceitful. I may not be perfect. But most posters have said I've done a pretty good job at this.

However, at almost every turn, you have misrepresented my argument, belittled me with small insults like when you type-casted your opponents as "data-obsessives", and outright accused me of intentionally lying and deceiving others.

These are not constructive behaviors. If you continue to do these three things, the solution is simple: I report you, and the conversation ends. This conversation has a chance to be civil and constructive. I'd certainly prefer it civil and constructive, since some of your points are insightful. But it’s up to you.

[capfan33]

A Case for Robinson > Giannis

Data discussion:

See, this is just outright misuse. Most of this is not readily available, or at least not in equivalent form, for the playoffs in Robinson’s prime (which is his primary criticism to an extent that goes well beyond every other top player). In the regular season, PIPM is not equivalent for both, and we do not even have Giannis’s PIPM from the past three years anyway (his 2020 LEBRON is the highest in the database going back to 2010 and represents a significant jump from 2019). We do not have RAPM for Robinson’s prime.

This gets back to Point 3 in the previous post. This is how we do statistical analysis... we look at the data, then apply context/film analysis, then come to conclusions. Is started with the data. Then applied context. Of course if you just quote the data-section and ignore the context, it looks like there isn't context...

That said, your point that we don't have full data for Robinson is absolutely true! And you'll notice how often I say just that when discussing older players. I reference the limits of data for older players in that very post

A small correction: it’s not true that PIPM is not available for Giannis for the last 3 years. We have PIPM numbers for both regular season and playoffs in 2020, and we have partial PIPM numbers in 2021.
It's also note quite true that we don't have RAPM for Robinson's entire prime. We have full RAPM numbers for him from age 31, 32, and 33, and partial RAPM numbers when he was younger. I'd consider that his prime, though not peak. If 30's the cutoff for prime, then Jordan was not in his prime in 96/97/98 and LeBron wasn't in his prime in 16/17/18. If you really don't think that's prime... well, let's just agree to disagree haha

Of course, the problem here is not just one of “playoffs versus regular season” or “what metrics are available for what purpose or in what form”. It is also treating it as something of a given that these reflect innate player quality -- but not consistently, mind you, or else we would need to start having some interesting conversations about 2003 Dirk (we will come back to that one) and 2016 Draymond and 2015/2017 Chris Paul and 2021/22 Jokic and oh right 2021 Gobert.

You suggest stats don't reflect innate player quality. Perhaps you mean what I've been calling "goodness"? Because by definition they measure value. You've called data driven approach "hypothetical", but it's just as hypothetical as averaging 30 ppg is. To me and to most people, you can count things, and when you get to the final number, you can say that that factually happened. That's not hypothetical.

As someone who pushes back somewhat against impact metrics as being the holy-grail they are sometimes espoused to be, this caught my attention. I evaluate players first and foremost based on the perceived goodness/skillset a player has, and use impact data to make sure I'm not missing anything major and to keep me honest. A data-driven approach by definition is empirical, not hypothetical, the way I approach player evaluation is much more hypothetical. So I'm not really sure where this comes from, because it's the opposite.

Impact stats are by definition empirical, how you interpret those statistics and what they may mean in a different situation is hypothetical. I'm not sure if I missed something in the previous exchanges, but as someone who comes at player evaluation from a fundamentally different angle to Draymond I 100% agree with him.

[trex_8063]

1. '95 David Robinson ('96 DRob, '94 DRob)
Not sure the best year to go with; each has a slightly differing selling point, and honestly I think he was near the same level in all three years.
I imagine this will be a contraversial pick, but it sort of follows in the same vein as my Kevin Garnett pick: two-way dynamos.
I think people sometimes fail to acknowledge how extraordinary an OFFENSIVE player Robinson was in the regular season.......they just focus on the playoff decline.
I've made this [only slightly hyperbolic] statement before: Robinson was essentially asked/expected to be Bill Russell on defense AND Michael Jordan on offense for those early-mid 90s Spurs teams.

And the crazy as **** part is: he was mostly successful during the rs.

Frankly if he HAD been able to maintain that in the post-season, he'd have more than a puncher's chance of taking the #1 greatest peak of all-time.

In addition to his off-the-chart box-derived metrics [in the rs], look at his rs impact metrics over this same three-year span:
'94 Top 5 APM (from colts18)
1. David Robinson: +7.31
2. Kevin Willis: +5.44
3. Karl Malone: +5.37
4. Hakeem Olajuwon: +5.10
5. Nate McMillan: +4.85
(NOTE: the separation between Robinson [at #1] and #2 is greater than the separation between #2 and #11).

'94 Top 5 AuPM (Backpicks)
1. David Robinson: +6.7
2. Karl Malone: +5.2
3. Nate McMillan: +4.8
4. Hakeem Olajuwon: +4.5
5. Kevin Willis: +4.3
(NOTE: the separation between Robinson [at #1] and #2 is greater than the separation between #2 and #6).


'95 Top 5 APM
1. David Robinson: +7.42
2. Shaquille O'Neal: +5.80
3. Karl Malone: +4.93
4. Anfernee Hardaway: +4.68
5. Scottie Pippen: +4.63
(NOTE: the separation between Robinson [at #1] and #2 is greater than the separation between #2 and #6).

'95 Top 5 AuPM
1. David Robinson: +8.7
2. Scottie Pippen: +5.9
3. Shaquille O'Neal: +5.6
4. Anfernee Hardaway: +5.6
5. Karl Malone: +5.3
(NOTE: the separation between Robinson [at #1] and #2 is equal to the separation between #2 and #24).


'96 Top 5 APM
1. Michael Jordan: +6.67
2. David Robinson: +5.89
3. Anfernee Hardaway: +5.26
4. Scottie Pippen: +4.99
5. Karl Malone: +4.89

'96 Top 5 AuPM
1. David Robinson: +6.7
2. Michael Jordan: +6.5
3. Scottie Pippen: +5.8
4. Anfernee Hardaway: +5.7
5. Karl Malone: +5.2


I mean, holy cow. You look at these three years, and he's not just outperforming the competition......he's obliterating it! Except for a prime Michael Jordan [in '96], there's no one remotely close to him in the rs.
And these are almost exclusively all-timers he's obliterating.
In the playoffs, he comes back to Earth: down to maybe being only maybe like the 2nd or 3rd-best player in the league. Oh dear.

So how much deduction should he get for going from GOAT-candidate in rs to circa-2nd best in any given year in the playoffs?
idk.....decide for yourself. But don't sleep on how friggin' unearthly he was in the rs.
For myself, he's near the region of Duncan/Hakeem/Garnett; usually just behind, but really really close.


2. '21??? Giannis Antetokounmpo ('20, '22)
Continuing with the two-way juggernauts: Giannis is definitely that. And because in some ways they're similar players, it feels appropriate to have Giannis somewhere within 1-2 places of David Robinson [even if it's Giannis you have ahead].
I reserve the right to re-order these three, btw; but for now I'll go with this order.


3. '22 Nikola Jokic ('21)
No one else immediately on my radar until these guys are in.
Jokic is one of the most complete offensive players of all-time, imo: Very efficient and very difficult to stop in low-post isolation? Check. Elite mid-range shot? Check. Good 3pt shooting? Check. Good FT shooting? Check. Double-team at your own peril? Double-check. Creation off the dribble or in transition [unusual in a big-man]? Check. Relevant offensive rebounding? Check. Reasonably efficient turnover economy? Check.

His offensive impact on these Denver squads has been off the charts, in an historic sense. While Murray was out, they were approximately the worst offensive team in the league when Jokic sat; and then when he was in the game, they were approximately the BEST offensive team in the league. I mean seriously: from last to first. Who the hell does that?

I also wonder if his D gets underrated at times too (at least when people label him a "weak" or "bad" defender).
The Nuggets were pretty much exactly league-average defensively last year; if your center [the most important defensive position]---who also played more minutes than anyone on the team---is "bad" or "weak", there must be others who are lifting your defense back up to mediocrity. Let's look at who that might be....

Aaron Gordon is a decent defensive forward to my eye, though not All-D tier or anything [imo].
Campazzo [8th in minutes] is pesky as hell, though also undersized [and a bit foul-prone]; I'm still inclined to think he's a plus defender overall, but not really moving the needle far (especially being only 8th in minutes played).
JaMychal Green [9th in minutes] at least contributes a respectable amount of DRebs/stl/blk, so maybe he's decent(ish) defensively? I can't claim enough eye-test on him recently to say. Some other guys aren't bad, but with the departure of Gary Harris their defensive backcourt is certainly nothing special (couple guys sort of medium-far down on the minutes list could be called legit BAD defenders).
So is this enough to off-set a truly "bad" defensive C (who also leads the team in minutes)?
idk, I'm a pinch skeptical anyway.

I'd also look at WHERE their defense performs well, and where it does not.
Where does it NOT perform well?.....
Opp TOV%: 26th in the league. This is generally something that is more predicated on your perimeter defense (they will usually be the ones generating turnovers or otherwise forcing errors thru ball-pressure). Not an aspect of defense we can reasonably expect a C to carry; imo, this is something that reflects more poorly on the perimeter defense. And fwiw, Jokic leads the team [comfortably] in spg, and is 2nd [to only Campazzo] in stl/100 of guys in their regular rotation; so he's a notable component of what few turnovers they do generate.

Opp eFG%: 20th in the league (and in particular: 28th in opp 2pt%). OK, this one DOES reflect poorly on Jokic. Certainly it depends to a degree on schemes and team defense, but this is certainly something you'd expect a good rim-protecting C to put a serious dent in. fwiw, Jokic leads the team in bpg [is 2nd to only Cousins in blk/100].

Opp FT rate: This one is a bit of a mixed bag as to who [which position(s)] has the bigger role. Really it's something of a total team effort/coordination, imo. They were 12th in the league in this category (respectable).

DREB%: This is perhaps the category a C can leave the most imprint on. They're on the interior, they are [or should be] boxing out; they're the ones securing the lion's share of defensive boards on most teams.
Denver's rank? 6th. It's the one defensive FF they were actually borderline-elite at.
Jokic was a close 2nd in the league in individual DREB% (behind only Rudy Gobert, and well ahead of 3rd place), and is far and away leading the Nuggets in defensive boards; there isn't anyone else on the team who even has 40% as many.

And when I watch Jokic, I see a guy with a reasonably decent defensive IQ, and passable effort. He lacks good lateral quickness or recovery speed, he lacks explosive leaping ability, he doesn't have Gobert's length, his rim protecting positioning [mostly with where he has his arms, imo] could be a little better. But his awareness and footwork......those are pretty good, imo.

So overall, idk......I just think his defensive short-comings are overstated sometimes.

[trex_8063]

It's also note quite true that we don't have RAPM for Robinson's entire prime. We have full RAPM numbers for him from age 31, 32, and 33, and partial RAPM numbers when he was younger. I'd consider that his prime, though not peak. If 30's the cutoff for prime, then Jordan was not in his prime in 96/97/98 and LeBron wasn't in his prime in 16/17/18. If you really don't think that's prime... well, let's just agree to disagree haha

I'm not in on the full context of this back and forth [and likely won't get fully up to speed], but I did want to jump in on this statement....

I would say it's inappropriate to define a player's prime based on age.
Was Bill Walton in his prime '79-'85? After all, he was only ages 26-32 in those years.

Everybody doesn't follow the same career arc.

Some guys are out of their prime by their mid-late 20s (Walton [injury]). Someone like Marvin Barnes was pretty much out of his prime by age 24 [drugs].
Guys like LeBron or Karl Malone are unicorns: still in their prime at age 36 (or arguably even at 37-38 in LeBron's case).
And then there's everything in between.

You can't necessarily say Robinson was still in his prime at age 31-33 just because some other guys were. It just doesn't work that way.


imo, you could argue that '98 (and maybe '99) were "late-prime"/"near-prime"/"extended prime" type years. Solidly in his prime? No, not really. And '97, definitely not (6 games playing limited minutes while injured, before finally conceding the season and undergoing back surgery is NOT prime-level).

[AEnigma]

Sorry to hear you're frustrated! Just to start: if you're frustrated by someone arguing for one player over another... you may be frustrated quite often around here. That's what we do. We discuss and we debate. And we don't always agree. Are you really so surprised by this?

Yeah that is surprising, because you seem a little more interested in selectively posting raw impact rankings and challenging everyone to follow them.

Rearranging this in and placing it at the top in light of your complaints about “tone”. I copied this over on a Google Doc though, so going forward I might miss some of the <crazy> emojis.

This seems like the crux of our disagreement (and it pops up multiple times in the rest of your post), so let’s start here.

Consistency of the Data-driven approach: Your primary argument that a data-focus is inconsistent replies on three things (at least as I understand it):
1) You suggest a data focus isn’t an “immediate baseline”. You call it a “hypothetical achievement”, compared to the achievement-focus (e.g. winning a championship) which is a “real achievement”
For 1: I’m not sure why being an “immediate baseline” should necessarily be proof that an achievement focus is consistent, while a data focus isn’t. It’s not true that having a good impact metric is a “hypothetical achievement”. It is a true fact that Giannis was an NBA champion in 2021. Likewise, it is a true fact that Giannis scored 28.1 ppg in the regular season. Likewise, it is a true fact that Giannis scored a +5.03 Goldstein RAPM. There’s nothing hypothetical about any of this.
Takeaway: Data not necessarily inconsistent.

2) You suggest a data focus is “convoluted,” while an achievement-focus is less complex.
For 2: You talk about increased complexity of data over “achievements” like this makes data inherently inconsistent. I’m not sure this is true. Things can be consistent and complex, or simple and inconsistent.
Takeaway: Data not necessarily inconsistent.

This is an incredibly strained reading and really sets a sour note for how you intend to approach this entire discussion. Honestly, this is like 90% of the reason I had no interest in participating in the top twelve: specifically because this is how far, far, far too many Curry fans argue, even the ones who play up some superficial politeness. I think you are being deliberately absurd by equating winning a title with scoring 28 points per game or having x RAPM “because all are facts”. Real “the trophy is just a hunk of metal” energy, although I recognise that would suit an approach prioritising “value to team” above all else. I do not believe that you think winning a title is just some random disconnected fact with the same real value as any other random metric, be it points or RAPM or rebounds or blocks or minutes played or literally anything. And if you did actually believe that, you would not waste so much time highlighting apparently arbitrary data achievements. I am not sure why you thought this was a good way to start, but I do not appreciate this type of time wasting.

To state what for everyone else should be the obvious: no, players are not setting out to maximise their RAPM, they are setting out to win a title. Playing as well as you can might correlate to RAPM, but you are not going to sabotage your backup in order to boost your own value. And to whatever extent this is correlated, it is even less correlated in the postseason, where losses and wins are much more significant and players are consistently more committed to doing what they can to avoid losses and earn wins.

And yes, it is inconsistent because there is no equivalent end target season-to-season. That was the point you somehow ignored. A player having a higher “impact” value to their team twenty years earlier does not equate to anything comparatively apart from what their value was to that specific team in that year. You say you understand this, yet you make it a key part of every comparison anyway. If you want to take an in-season ranking approach, alright, fair enough. David Robinson was top impact in 1994 and 1995. Giannis was not top impact in 2021 (Jokic). He was in 2020 though, so why are we throwing that out? Well, standard deviations maybe. This again is not really what basketball is about, but hey, it is an approach. But the problem with that approach is that it inherently asks us to look backward at leagues where the best players were more substantially separated from their competition, and seeing as you do not really seem committed to doing that, where does that leave us. If your real intent was just to really lean into the idea of making your approach look excessively convoluted, mission accomplished I guess, but you did not really do anything to work against the idea that your approach is inherently inconsistent when compared to someone going, “Wow, they won a title as a best player with an impressively valuable performance, good stuff.”

3) You suggest a data focus lacks sufficient context and analysis (e.g. when comparing players in different eras, different situations, potentially with different teammates/opponents).
For 3: This is the most compelling point. It’s true that just looking at stats without context can be misleading for the reasons you suggest. But here’s the problem: this is a bit of a straw-man. This implies that everyone who used data in the past threads didn’t provide context.

But we did! I provided plenty of context… Now, you may not agree with someone’s evaluation of the context. But the fact that you disagree with someones interpretation of the context =/= no context was provided.

If I say David Robinson was on a 20-win team, but in fairness to him, that 20-win team fired their coach midseason and was trying to tank for a generational rookie, do you think that is worthwhile context. If you rely on selective context to make your argument, what does that say for how much weight anyone should place in your argument.

Regardless, this is a lazy dodge. Yes, I think you did a bad job interpreting the context, and I went through why and how.

I compared eras,

You compared eras by saying Giannis has spacing now that he would not have in the 1990s. Good job.

I analyzed the potential biases of different stats,

You mean like saying BPM probably underrates Robinson’s defence? For the most part you are just posting, “Robinson wins in x number of stats.” That is not analysis, and you are openly stating that being higher in a stat means you “win”. Here again, I do not think you are getting the point.

compared the team’s fits,

While ignoring how Robinson’s “poor fit” juiced his impact.

I compared performance against better defenses,

Where in that post did you go through Robinson’s performances against good defences compared to Giannis’s.

I compared their different skillsets…

Partially, to the extent that you wanted to favour Robinson. You said Robinson was more scalable, that he was a better rim protector, that you liked his help defence (which is not really a comparison seeing as Giannis also gets most of his value through help defence), and that you liked his passing more. You also said that you think Giannis’s defence would still hold up in the 1990s, but that lack of spacing and stricter dribbling would hurt him, and I pointed out where that comparison fell short.

It’s worth noting — many of the times I provided context, you used this as evidence that I was biased: you say I was “throwing out a year that is inconvenient to the narrative [I] want to push or that the metrics aren’t perfect.”, when really I was providing context for Robinson.

Well, no, you excluded context lol. You cut out how badly his scoring dipped in 1998 even though he had a good team and a top tier partner. And for Giannis, you cut out 2018 in assessing how he “always” dropped off in the postseason.

Similarly, you suggest that many times knowledgeable analysts provide context to the data, they’re actually just showing inconsistency.

Love the framing of “knowledgeable analysts”. No, the point is that data is not really worth blindly relying upon. Yet consistently with you we see you starting from the premise that x player “wins” in direct data comparisons -- which again is not how impact assessments work, and is not changed by making selectively arranged “arguments” for those players to justify using the data disparity as a real ranking.

For example, you suggest that it’s inconsistent that the data-driven approach considers 2010/2011 Dirk > 2006/2003 Dirk. You make simalar arguments with mentions of Draymond/Paul/Gobert. But… taking 2011 Dirk >2006/2003 Dirk is the very thing that context/film-analysis would suggest! Taking Draymond/Paul/Gobert lower is the very thing that context would suggest!

Sure, but you are not doing that. Why are you taking 1994 Robinson over 1995 or 1996 Robinson.

It’s also what an achievement-approach would suggest, which is the one you’re a proponent of.

If I were focused on achievements I would be way higher on Moses and Shaq and way lower on Garnett. My focus is mostly on playoff impact with a preference toward two-way bigs.

So… from your perspective, either data-driven approach either doesn’t consider context but should (which isn’t true) or they do consider context which shows inconsistency with data. This logic seems a bit inconsistent, no? Either we should consider context or we shouldn’t. In truth, we do. But just because you personally disagree with someone’s analysis of context doesn’t make it invalid or inconsistent.
Takeaway: Data not necessarily inconsistent.

This is really funny coming from the person who in every thread starts with mock confusion at how no one seems to agree with what the data oh so very clearly says.

Inconsistency of the “Achievement”-approach: Instead, I’d argue taking an “achievement focus” opens you up to more potential biases and inconsistencies. People have discussed “winning bias”, which you mentioned too, where the results of the team against a certain opposition are conflated with the greatness of the individual.

Who in the top eleven do you feel suffered from winning bias.

I’ve also raised the idea of a “scoring bias” (overvaluing scoring and undervaluing playmaking), and “athleticism bias” (overrating athleticism and underrating skill/BBIQ),

Neither of these are really points to Robinson’s advantage, especially with Walton in the discussion. Scoring bias is not a point to Curry, although I recognise he is more the target of the other.

An achievement-based approach would open you up to all of these. Data doesn’t care whether you won the championship or lost… it just tells you how much you helped your team by. Data doesn’t care about whether your impact comes from scoring or passing. Data doesn’t care whether your greatness comes from athleticism, skill, or BBIQ. Data doesn’t care about a single memorable highlight which (in the vast majority of cases) doesn’t actually change who helps their team win more.

Impacting your team more does not mean you are a better player. That is the entire point. You keep saying you understand that and then turning around and writing stuff like this.

Could an achievement-based approach lead you astray? Let’s try an experiment. Let’s do an achievement-based approach for Garnett. You suggest that the “pretty wise range of possibility” in Garnett’s team results suggest a wider range of possibility compared to the consistency of “guys like Duncan and Magic and Bird [who] elevated their games against the best competition.”
Well, if team results are what we’re going by, that would suggest Garnett is better in 2008 vs 2003/2004, right? I’d be hard pressed to find anyone on this board who would agree with that. So how do we know that Garnett was actually better in 04, using the “achievement-based” approach? Well, we’d have to look at context…. the same context people using the data-driven approach are providing. Except where just looking at “achievements” [of the team] gives us no way to isolate how much the individual helped the team, just looking at the data gives us a very precise ability to isolate how much the individual helped the team.

Garnett was not a better player in 2004 because the data said he was more impactful. Once again, you openly fail to understand things you earlier claimed you understood.

You do not see any “achievement-based” approaches preferring 2008 Garnett because they do not see that as meaningful an achievement. The better example here would be with those select few who took 1971 Kareem over 1974 or 1977 or even 1972 (maybe 1980 too?) Kareem. But those too were an extreme minority even among the achievement group. If you are that uninterested in figuring out why people vote the way they do, how do you expect to convince people to change votes.

In sum: both methods need context. The data-driven approach gives us an much better insight into how much the individual was actually contributing to wins, while avoiding biases like winning bias, scoring bias, athleticism bias, and big moment bias. An achievement-based approach… doesn’t tell us how much the individual helped their team win and opens us up to those biases far more.

And why is a “value to team” bias better. Or perhaps you would find that question more valuable if I peppered you with “Takeway: value to team bias not necessarily better”.

You say that I should "spend much time truly trying to understand how other people are making their comparisons." Is this not a two-way street? The burden's just on me and on this side, and not on the other side?

You have both made it very clear how you are making your comparisons and also are the person most loudly unclear why people have not adapted their vote to match yours, so yes.

There's a pretty big difference between disagreeing with an evaluation and saying they are biased and aren't trying to understand a different perspective.

True, and the difference is that most people who say they want to figure out why people are not voting a certain way make an actual effort to think about what those differences are rather than taking it as a given that their point is the best and most accurate way to start.

This gets back to Point 3 in the previous post. This is how we do statistical analysis... we look at the data, then apply context/film analysis, then come to conclusions. Is started with the data. Then applied context. Of course if you just quote the data-section and ignore the context, it looks like there isn't context…

But your “context” is not an analysis of the data. That is the entire problem. You are treating the “context” section like an area where you assuage any concerns with what the data says rather than looking at why the data says what it says or whether you should even be using them comparatively to begin with.

That said, your point that we don't have full data for Robinson is absolutely true! And you'll notice how often I say just that when discussing older players. I reference the limits of data for older players in that very post.

You reference them but do not really do anything with them, because functionally you are more interested in what you think existing data says.

A small correction: it’s not true that PIPM is not available for Giannis for the last 3 years. We have PIPM numbers for both regular season and playoffs in 2020, and we have partial PIPM numbers in 2021.

True, I was off by a year, and his 2020 PIPM is indeed the third highest on record after 2009 Lebron and 2016 Draymond.

It's also note quite true that we don't have RAPM for Robinson's entire prime. We have full RAPM numbers for him from age 31, 32, and 33, and partial RAPM numbers when he was younger. I'd consider that his prime, though not peak. If 30's the cutoff for prime, then Jordan was not in his prime in 96/97/98 and LeBron wasn't in his prime in 16/17/18. If you really don't think that's prime... well, let's just agree to disagree haha

… Do you think primes are tied to strict age ranges?
Maybe you can make the case for 1998. I would sort that as an approximate equivalent to a 2018 Chris Paul season: step down from what I would consider his actual prime, but close enough. From then on you are getting more like 2020/2021 Chris Paul, which is still good but distinctly not his prime, no.

You suggest stats don't reflect innate player quality. Perhaps you mean what I've been calling "goodness"? Because by definition they measure value. You've called data driven approach "hypothetical", but it's just as hypothetical as averaging 30 ppg is. To me and to most people, you can count things, and when you get to the final number, you can say that that factually happened. That's not hypothetical. But yes, I absolutely agree, context needs to be provided from there! Value in a given role is not the same as goodness in any role, and we need context to decipher this. A player can be super valuable in a niche role/fit, but not as good as their value suggests if forced to play in another situation. This is the kind of context you accuse me of not applying... but it's also the same context that I do apply to understand why someone like Gobert shouldn't be the 13th GOAT Peak. I'm also not sure the data does indeed support all the players you mentioned being in this tier. To me, that misrepresents the data.

This is a fine answer and more what I was expecting (took your time getting here), but it does not really explain why you are so attached to 1994 Robinson specifically.

Paul: I provided a statistical comparison with Chris Paul in a previous thread which showed across the majority of metrics, he’s a hair below the modern players we’re considering. So it’s perfectly consistent to not include Paul at #13 in a data-driven peak list… when the data doesn’t support him being #13 in a data-driven peak list.

How early would you vote for Chris Paul. Would you vote for him over Kobe? Would you vote for him over 2009 Wade?

Jokic: You suggest that I’m hypocritical for not discussing Jokic over Giannis. Except… I have. In the very post you’re quoting, I said that Jokic is in the same tier statistically as Giannis, and that I would personally prefer peak Jokic to Giannis. This seems like a bit of a straw-man, no?

Is he in the same tier in the postseason?

Gobert: To my knowledge, Gobert is also lower across the spectrum of metrics, so in a data-driven list, it would also be consistent to have him lower. Perhaps I’m missing some metrics you’re referring to?

I am mostly referring to the idea of role importance overstating a player’s “goodness”, but yes there were a fair number of metrics that had Gobert top two in the league along with Jokic in 2021, and given that we are talking about Jokic potentially as early as #15, that stands out. I am not sure how many Kobe seasons can claim the same value as regular season Gobert, for example.

Draymond: Here, we face two issues.
First, sample size. 2016 Draymond shoots to the Top Tier of plenty of stats. But other Draymond years don’t, and the drop-off is far larger than the drop-off for other top single-year peaks we’re considering. I’ve argued that we should weigh single-year peaks heavily (particularly when there are contextual factors that might explain a drop on either side), but that we should still consider ~3-year peaks to get a larger, more consistent sample.
Second, is how to incorporate context into impact metrics. 2016 Draymond provides a ton of value, but when applying context, basically everyone agrees it's from having an outlier shooting season and outlier fit, which doesn't suggest how truly good he'd be in other roles. See above / previous post for more on how to include context in data.

But Draymond is consistently exhibiting high impact in the postseason. You liked to use minutes with only Curry to show that based on that sample Curry could be great without his teammates. But you can do something similar with Draymond over the years, so is it really all about fit? I mean, yeah obviously Draymond has next to no regular season floor-raising value (2020 made that clear), and that is immediately discrediting for a lot of people, but since you are so committed to the data, are you as confident in saying he would not be good in the postseason specifically on other teams?

Interestingly, this was an unexplored issue I had with Boston Garnett (not that it is overly relevant to assessing 2003/04 Garnett): he put up giant numbers with them but without the results were more middling (not bad, but well short of Curry or Lebron).

This seems somewhat testy in tone for what is yet another straw man. I didn’t mention 2018 Giannis because, at least to my knowledge, not a single other poster even mentioned 2018 Giannis, much less were arguing that his 2018 performance should be indicative of his 2021 impact. Look, I'm happy to have a discussion about 2018 Giannis, but when you say something I disagree with (which is clearly explainable from the context of the rest of the discussion), my immediate instinct isn't to accuse you of intentional omission or deception. I'd strongly suggest you avoid such accusations in the future.

Do you think he became innately less resilient from 2018? You explicitly said that he declined in every single series of his prime. That is an outright misrepresentation unless you exclude 2018 from his prime, and there is no logical reason to make that exclusion. Chalk it up to poor phrasing, fair enough, but when it is part of a consistent pattern of underselling players who do not have the giant data footprint you want, hard to treat it as just an innocent instance of misspeaking.

I appreciate the emoji though!

The recipe isn't quite "giant data" = "best player". It's "better data" + "better context that suggests the data isn't too biased towards the statistical winner / against the loser" = "better player".

… But it is biased. Being relied on more does not make you better. That is most of what this is saying at the top level. I am not arguing data should be thrown out completely, but when you use it as your starting point, that is ultimately what you are doing when you take the approach that we should default to the bigger data footprint barring strong alternative suggestions, and especially when the priority seems to be regular season production in particular.

2009 has better impact, but the context/film analysis is worse. That's why people who support the data-driven approach favor the older LeBron. We're applying the very context that you suggest we're missing.

So why not do that for 1994 Robinson?

Of course it's not enough to look just at relative drop-offs, and I never meant to suggest it was. I actually suggest otherwise at some point in the past few threads. It's great than Giannis is the better scorer. Outright. But what if Robinson/Garnett are the better playmakers or off-ball players or defenders? Well, then the primary advantage Giannis has over them is in scoring.
And what if Giannis' scoring decline is large against good opponents in the playoffs? Well, then his biggest advantage over Robinson/Garnett declines. So if Giannis doesn't shrink the defensive/playmaking/off-ball/etc. gap in the playoffs (which has been debated but certainly not proven), then Giannis might just end up being less resilient and less good overall than Robinson/Garnett. Outright.

But you did not conduct any real analysis of what would make Robinson a better playmaker (distinct from passer, which itself is already pretty contentious). You barely touched on defence, and gave Robinson the advantage by glossing over his weaknesses. And for postseason scoring, the only area where Robinson does not decline similarly or worse is against bad defences (calling back to that lessened reliance on free throws), or I guess in a scenario where he does not need to shoulder a large scoring load (which would reduce his data footprint anyway). That is not exactly nothing, but it is a lot less meaningful when looking at the postseason as a whole.

First off, once again this seems like a straw man / misrepresentation: I didn't throw out "a year" to intentionally omit for Giannis, as I mentioned above, nor is this just "a year" for Robinson. I said "1999 onward", which to me suggests multiple years.

But why start in 1999. The team was good in 1998, although they certainly missed Sean Elliott against the Jazz. Again, looks like a broader pattern of cherrypicking.

And again with the resilience with Duncan comment, it seems like you missed the point I was making. Perhaps I could have been clearer? Let me say unequivocally: Of course Robinson's resilience didn't magically get better with Duncan! So... why did he stop dropping in the playoffs?

… He did not. I said that explicitly. I got the point you were making, but you completely read past mine. He continued to drop in the playoffs with Duncan, although when you exclude 1998 for arbitrary reasons, that drop does look like less.

Well, he fit better as a second option.

Because he is not a good scorer.

There's nothin magical about rookie Timmy.. the reason Robinson was better was because he was playing a role he fit more.

A role that asks him to score less and thus provide less raw value.

And while it's of course true we can't see Giannis play with Duncan, we can infer how he might play as a second option based on his actual skillset and how he meshes with other offensive players on his team.

Why would Giannis be the second option?

You are again doing a funny thing here where you are assigning arbitrary penalties. Robinson gets rewarded for being a worse scorer and thus accommodating Duncan (a better scorer). But there was a lot of consternation over how limited Duncan is on offence too. Why would it not benefit Duncan to be a second option to Giannis the way it benefits Robinson to be a second option to Duncan? And then in the playoffs, why would it not be better to have resilient scorer Tim Duncan getting secondary attention (or otherwise alleviating the attention on Giannis) versus consistent playoff shrinking David Robinson?

Let me summarize my argument, since that might be more productive than over-debating points that seem like straw-man arguments of mine.
Yes, Giannis is the better scorer and better offensive first option, which is a point for Giannis > Robinson.
But: Robinson may have other offensive skills over Giannis, so it's not a strong point for Giannis.
But: neither a great offensive first options compared to the rest of the competition, so it's again not a strong point for Giannis.
But: Robinson is the better defender. Point for Robinson
But: Robinson is the far better offensive second option than Giannis. Point for Robinson.
Thus: Robinson > Giannis.

…
I mean, to be blunt, those are bad arguments. I am not sure where you keep seeing straw-men when no I was very certain these were your arguments. I think you might be conflating me saying these are weak arguments with saying they are weak because I am relying on straw-men, because you seem to be so confident in what the data tells you that it is impossible for you to visualise disagreement. Which has been my core point throughout.

Starting from the bottom, why is this a point. Or rather, how is this a point. The premise is that being a bad scorer means that you lose less value as your scoring responsibility decreases. Fair enough. But then… why would we stop at Robinson? Why not extend to Walton. Why would we not extend that logic to pretty much every volume scorer. Hell, why are we stopping at second option. This is again essentially praising someone for a limitation, and on top of that, it is a comprehensively baseless one: why would an offence that somehow manages to make Giannis into a second option be worse? Giannis already has massive gravity, and apparently he is being paired with a guy who attracts even more attention or is otherwise even more capable of scoring when provided soft coverage? Sign me up, that sounds like a recipe for best offence in the league. And in the alternative, where he is simply paired with the type of player who would make David Robinson take a backseat? Again, would love to see it, because that player would probably be a first option on at least half the other teams in the league.

Related to that point, no, Giannis is a legitimately great scorer. Even with his postseason dropoff, very few players can claim his level of scoring volume and efficiency. Do I trust him as much as Dirk or Jokic? No, but you are still getting a legitimately elite scoring anchor, and Dirk/Jokic do not need to be the bar in that comparison. Like, man, the more I think about this, the more ridiculous it becomes. Kyrie is a better scorer than Giannis. Kyrie gets more of his value from scoring than Giannis. I am very concerned about Kyrie playing next to Kevin Durant oh look at that the offence just became historically great. The immediate counterpoint of course is that Kyrie is portable. Okay, fair enough. Maybe we use James Harden then. Oh look at that, the offence is still incredible, probably good enough to win a title if Harden does not have some wonky hamstring or something. This criticism only works if we are specially pairing Giannis with… uh… Shaq, maybe? But hell, good luck guarding a frontcourt of Shaq and Giannis. Dwight Howard or Shawn Kemp would be pretty bad fits I guess, but they would be bad fits with Robinson too lol (not as bad of course) and both of them would still be the second options by a clear margin.

This is not a coherent complaint. More natural second option does not mean better second option. Duncan + Robinson did not really make for a notable offence. It was just more reliable than leaving either of them on their own, and of course combined two top tier defenders (which you are getting with Giannis too). Speaking of which, defensively you have done basically nothing to prove Robinson’s advantage, and in any case, you have certainly not done close to enough to prove that the disparity in their defence is larger than the disparity in their scoring / playmaking / general offensive value. Like, even if you think this is a +2 offence and +4 defence situation, there is nothing suggesting that the defensive gap -- if one is even present -- is proportionally larger than the gap in their offence to the extent it would need to be for this to make sense.

This again seems like somewhat of a straw-man.

The whole concern is whether Giannis' poor playmaking is the thing that's limiting his teammates' shooting. It's great that both Curry and Giannis draw doubles. Legitimately, it's great! But I showed in film analysis that these double teams on Curry led to easier shots for his teammates, e.g. off of Curry's superior passing or superior off-ball movement.

If Giannis draws the double with the ball but struggles to pass... then his team might not get the more efficient shot, which is the whole point of drawing the double team! If Giannis draws the double without the ball, but his off-ball movement isn't in a way that leads to easier shots for his teammates... then the double team doesn't really help the offense.

Like I say in that very post, these are concerns, not proven. But they also haven't been disproven in any in-depth film analysis either. These are all things I say in my post / in previous posts that I link to in that post. You mention you're quite frustrated. Well.. I'm also frustrated too, since my points keep getting straw-manned!

I wonder, does Robinson struggle against elite offences because he is actually not that valuable a defender? I mean, I have not seen anyone do in-depth film analysis showing that is not the case. I mean, it is weird, right? It is just a concern, but you know, that is what great defenders are supposed to do, right? If his help defence is so good, why could he not help stop the other team? Just asking questions here.

I think it's time I address tone.

When I think I’ve misunderstood another poster’s meaning or intention, I simply ask. When I think I understand another poster that I disagree with, I simply explain why I disagree.

To my best effort, I never straw-man another poster, belittle them, or accuse them of being deceitful. I may not be perfect. But most posters have said I've done a pretty good job at this.

However, at almost every turn, you have misrepresented my argument, belittled me with small insults like when you type-casted your opponents as "data-obsessives", and outright accused me of intentionally lying and deceiving others.

These are not constructive behaviors. If you continue to do these three things, the solution is simple: I report you, and the conversation ends. This conversation has a chance to be civil and constructive. It’s up to you.

Yeah civil and constructive tone is when I throw a bunch of crazy emojis at you and say maybe you are not suited for discourse here and say everything you write is a straw man if I do not like how it implies my argument is not as good as I want it to be and more broadly complain about how confusing it is that people are not just copying my vote even as I use excessively selective frameworks to support how I vote. Feel free to report me for not doing that.

[DraymondGold]

It's also note quite true that we don't have RAPM for Robinson's entire prime. We have full RAPM numbers for him from age 31, 32, and 33, and partial RAPM numbers when he was younger. I'd consider that his prime, though not peak. If 30's the cutoff for prime, then Jordan was not in his prime in 96/97/98 and LeBron wasn't in his prime in 16/17/18. If you really don't think that's prime... well, let's just agree to disagree haha

I'm not in on the full context of this back and forth [and likely won't get fully up to speed], but I did want to jump in on this statement....

I would say it's inappropriate to define a player's prime based on age.
Was Bill Walton in his prime '79-'85? After all, he was only ages 26-32 in those years.

Everybody doesn't follow the same career arc.

Some guys are out of their prime by their mid-late 20s (Walton [injury]). Someone like Marvin Barnes was pretty much out of his prime by age 24 [drugs].
Guys like LeBron or Karl Malone are unicorns: still in their prime at age 36 (or arguably even at 37-38 in LeBron's case).
And then there's everything in between.

You can't necessarily say Robinson was still in his prime at age 31-33 just because some other guys were. It just doesn't work that way.


imo, you could argue that '98 (and maybe '99) were "late-prime"/"near-prime"/"extended prime" type years. Solidly in his prime? No, not really. And '97, definitely not (6 games playing limited minutes while injured, before finally conceding the season and undergoing back surgery is NOT prime-level).

Quite right trex_8063, that age is not the sole marker of prime. It was used as a shorthand example to make a point, but perhaps I should have been clearer with my reasoning.

To do just that: most people have Robinson's 3-year peak as 94-96. I personally do. So I'm hard-pressed to think that one year after a peak, Robinson's immediately out of his prime. That would be quite the drop-off! Among all-time greats (you're right in non-all timers like Marvin Barnes, there's certainly quicker decline), I'm not sure we see that kind of immediate peak-to-past-his-prime drop-off ever happened, excepting injuries... unless I'm forgetting anyone? To me, Walton's clearly a drop caused by injury. I'd say your characterization as late prime/near-prime is perfectly accurate for 97/98/99. with each year being successfully further from true prime.

Sorry to hear you're frustrated! Just to start: if you're frustrated by someone arguing for one player over another... you may be frustrated quite often around here. That's what we do. We discuss and we debate. And we don't always agree. Are you really so surprised by this?

Yeah that is surprising, because you seem a little more interested in selectively posting raw impact rankings and challenging everyone to follow them.

I made a post like everyone else, provided statics in case people were interested, and invited discussion and debate. I haven't gotten the impression anyone felt personally challenged by my post or my opinion, but if anyone else was, please do let me know! I'm certainly open to being better as a poster!

...

I think it's time I address tone.

When I think I’ve misunderstood another poster’s meaning or intention, I simply ask. When I think I understand another poster that I disagree with, I simply explain why I disagree.

To my best effort, I never straw-man another poster, belittle them, or accuse them of being deceitful. I may not be perfect. But most posters have said I've done a pretty good job at this.

However, at almost every turn, you have misrepresented my argument, belittled me with small insults like when you type-casted your opponents as "data-obsessives", and outright accused me of intentionally lying and deceiving others.

These are not constructive behaviors. If you continue to do these three things, the solution is simple: I report you, and the conversation ends. This conversation has a chance to be civil and constructive. It’s up to you.

Yeah civil and constructive tone is when I throw a bunch of crazy emojis at you and say maybe you are not suited for discourse here and say everything you write is a straw man if I do not like how it implies my argument is not as good as I want it to be and more broadly complain about how confusing it is that people are not just copying my vote even as I use excessively selective frameworks to support how I vote. Feel free to report me for not doing that.

I used two crazy emojis when you first openly accused me of being a poster with bad intentions, then openly accused me of intentional deception and lying. If the crazy emojis were insightful or hurtful, I apologize.

However, it seems like you're still intending to use sarcasm in this conversation, and you stand by your statements that I have bad intentions and am actively deceiving others. I think the productive part of our discussion has past, which saddens me, but I don't have much interest in discussing further with someone who has not been willing to be civil with me, listen to my perspective, or give me the benefit of the doubt.

...

To other posters: If any part of AEnigma's description of my posts rings true to you, if it seems like I'm actively trying to deceive you or have bad intentions, again please don't hesitate to let me know! I'm absolutely open to being a better poster. (if you do feel this way, I'd love to hear it in a slightly a more constructive way without sarcasm or calling me a liar). For the most part, I've found posters here intelligent and kind, and I'm hoping this brief interruption doesn't sour the rest of the discussions.

[AEnigma]

Not being objective is not the same as being a liar.

To that point, I do not think you are lying when you talk about 1997 David Robinson, but it does call into question how objective you really can be about his career arc lol.

[falcolombardi]

Not being objective is not the same as being a liar.

To that point, I do not think you are lying when you talk about 1997 David Robinson, but it does call into question how objective you really can be about his career arc lol.

97 was robinson missed season

Did you mean 96 perhaps?

[AEnigma]

Mm, read the post directly above that

[Dutchball97]

If the priority is to vote for players with an MVP and Finals MVP, what puts Moses above Willis Reed or Bill Walton (or Mikan or Pettit, for the older school voters)? Why put so much stock into Moses winning a three game series and then going “fo’ [five] fo’” on a team that had been to two of the past three finals (and probably would have won a championship in the year they missed if they had made one more basket in Game 7)?

It's nice to see someone else push back against the "data-first" approach that many voters seem to employ. While I intensively try to use stats in my evaluations as well I agree it's more productive to first look at what actually happened and then check if the data backs it up instead of simply looking at your preferred stat and voting for the highest number. In most cases stats (whether raw, box or impact) aren't meant to be compared straight up year to year. I'm not trying to invalidate the posters who do use the "data-first" approach but sometimes it moves the discussion overwhelmingly to certain players, while discounting others, especially when relying on limited and flawed stats from before the +- era. Actually sounds a lot like the "winning bias" I'm acused of so maybe we're not that different after all.

I did want to touch on the quoted part of your comment though as I have Moses #2 on my ballot with him being the clear MVP and FMVP counting as an important feat for me. I skipped the whiteness debate because that's not really something that's been relevant for a while now where I'm from and I can't see across the ocean how much of the racial bias in the US actually impacted MVP voting.

As to the mentioned players I don't think their case is as strong as Moses from an achievement perspective. In 1970 Reed barely beat out West and Kareem for MVP and I'm not sure it was even completely deserved. The 1970 Knicks in general were a two-headed dragon with Reed and Frazier where all the credit went to the big man by default when it wasn't like that at all. You could make that argument for Moses and Dr J but Moses was undoubtedly the best player on his team in both the regular season and play-offs. Dr J was definitely also not someone who didn't get his fair share of credit. Walton as a player in a vacuum probably should've been voted in already but across a whole season I don't see him having as strong of a case. He won FMVP in 77 and then got better the year after and somehow won the 78 MVP while only playing 58 games before getting injured early in the post-season. If there was a way to mashup the 77 and 78 seasons he'd be top 10 for me but as it stands I still have a couple more guys over him. Mikan is tough because relatively to competition he is at worst a top 5 player in league history but the problem is his competition is also without a doubt by far the worst in NBA history. Pettit didn't have his strongest regular season in 58 (even though he was so consistent it isn't too far off his best outings) and then got somewhat upstaged by Hagan in the play-offs. 59 seems to be the pick for his peak here and he did win MVP then but lost right away to Baylor's Lakers in the post-season.

[f4p]

To other posters: If any part of AEnigma's description of my posts rings true to you, if it seems like I'm actively trying to deceive you or have bad intentions, again please don't hesitate to let me know! I'm absolutely open to being a better poster. (if you do feel this way, I'd love to hear it in a slightly a more constructive way without sarcasm or calling me a liar). For the most part, I've found posters here intelligent and kind, and I'm hoping this brief interruption doesn't sour the rest of the discussions.

AEnigma seems to be an irascible sort, for no apparent reason. I want to agree with a lot of what he is saying, but he makes it difficult. Obviously, I haven't agreed with a lot of your takes in this project, but the posts are detailed and well thought out and we all have our reasons we focus on for picking people and they won't always agree. And I personally like reading posts I don't agree with if they make me think about my approach.

[Dutchball97]

To other posters: If any part of AEnigma's description of my posts rings true to you, if it seems like I'm actively trying to deceive you or have bad intentions, again please don't hesitate to let me know! I'm absolutely open to being a better poster. (if you do feel this way, I'd love to hear it in a slightly a more constructive way without sarcasm or calling me a liar). For the most part, I've found posters here intelligent and kind, and I'm hoping this brief interruption doesn't sour the rest of the discussions.

It's great to see when someone puts so much time and effort into their evaluations as you're doing. Obviously we're going to disagree here more often than agree due to our different approaches but I don't think that makes either of our opinions more or less valid than the other.

As a small piece of constructive criticism, I did find quoting people asking for a clarification for their votes when they don't allign with your datasets to be a bit on the direct side. Not everyone is going to value the same stats in the same way so it might be more productive to address specific points others are making (which you've also already been doing) instead of trying to focus the discussion on the datasets you're presenting. Even then you remain polite throughout so it's really not that big of a deal, just a small thing that might help.

What sometimes gets a bit lost in the sauce because of things heating up is that these projects are meant to stimulate in depth discussions and a process of learning instead of coming up with an authoritative and exact ranking. Nobody is ever going to agree with the list as a whole due to so many people with different opinions coming together to create an aggregate ranking but if we all come out of it a bit more knowledgable then the project was a success. I've found that even if you completely disagree with someone you can still learn something from their arguments and maybe even move closer together on certain points.